Skip to Main Content

Marvin Goodfriend and the Zero Lower Bound

Honoring Marvin Goodfriend

Goodfriend, Marvin. 2000. "Overcoming the Zero Bound on Interest Rate Policy." Journal of Money, Credit and Banking 32, no. 4, part 2 (November): 1007-35.

The zero lower bound (ZLB) on short-term interest rates has become a central issue in contemporary monetary economics. Fundamentally, the ZLB exists because households and businesses can choose to hold cash, which pays zero nominal interest, rather than accept a negative return on their short-term investments.1 (Other considerations, including concerns about the possibly adverse effects of negative rates on the financial system, have also made some central banks unwilling to cut short-term rates below zero or very far below zero.) Together with long-term global declines in the so-called neutral rate of interest — the interest rate consistent with full employment and price stability — the ZLB has significantly reduced the ability of monetary policymakers to ease policy through traditional short-term interest rate cuts.

For example, the Fed's reaction to a typical recession before 2000 was to cut its target for the federal funds rate by 5 to 6 percentage points. Today, in contrast, the Fed's scope for rate cuts is probably at most 2 or 3 percentage points on average — even less in Europe or Japan. The reduction in policy "space" available through traditional methods has led central banks to experiment with alternative tools, including quantitative easing and detailed forward guidance. Central banks have also developed new policy frameworks aimed at mitigating the effects of the ZLB, such as the Fed's flexible average inflation targeting approach.

During the decades following World War II, the ZLB was treated by most economists as a Depression-era curiosity. However, it became relevant again in the 1990s, when the Bank of Japan — in the aftermath of the collapse of the country's stock and real estate markets — struggled with deflation and the lower bound with little success. Many at the time saw the Japanese economy as quite different from that of the United States. Its monetary institutions and practices differed from ours, and key structural features — including high saving rates, slow population growth, and a sharp slowdown in productivity — were conducive to a lower neutral interest rate. Nevertheless, the Federal Reserve was sufficiently interested in the implications of low rates of inflation and interest to organize an October 1999 conference in Woodstock, Vermont, on the subject of "Monetary Policy in a Low-Inflation Environment." Then an academic, I was fortunate to attend. It was a stimulating few days, full of untrammeled discussions and new and sometimes radical ideas. Our debates, in the formal sessions and informal get-togethers, plumbed deep issues in monetary theory. I suspect that most of the attendees enjoyed the blue-sky thinking at the conference but underestimated the practical significance that those discussions would have a few years later. I know that I did.

Marvin Goodfriend's paper at the Woodstock conference, "Overcoming the Zero Bound on Interest Rate Policy," was a highlight. Marvin's article — which would appear in 2000 in the Journal of Money, Credit and Banking — anticipated a number of the approaches that central banks would use (or contemplate) during and after the global financial crisis. Like Irving Fisher, John Maynard Keynes, Silvio Gesell, and other great economists who had thought about these issues in earlier eras, Marvin did not hesitate to recommend radical institutional reforms — notably changes that would eliminate the constraint on monetary policy posed by the availability of currency. He would return to these issues in a number of subsequent writings, including in a paper presented at the August 2016 Jackson Hole conference, entitled "The Case for Unencumbering Interest Rate Policy at the Zero Bound." Rereading these papers today, I am impressed by their prescience and insight. I am also struck by Marvin's conviction that his job was to get the economic analysis right, without concern about the politics.

A general approach to overcoming the ZLB, favored by some economists, is to try to manage inflation or inflation expectations directly, through forward guidance or by announcing changes to the central bank's policy framework. For example, raising the central bank's medium-term inflation target — if fully credible — should also increase the neutral nominal interest rate and thus increase policy space. As Marvin pointed out in his 2000 paper, though, even putting aside issues of credibility, this tactic is inefficient. Raising the target (and assuming the new target can be reached) imposes higher inflation at all times but benefits the economy only at times when the ZLB would otherwise bind. Moreover, it reduces but does not eliminate the adverse effects of the ZLB in severe recessions, in which deeply negative real interest rates are needed. Alternative policy frameworks that are at least theoretically more efficient, though also more complex, depend on the central bank's ability to vary inflation and inflation expectations according to the state of the economy, for example, through price-level targeting (Wolman, 1998; Krugman, 1998; Eggertsson and Woodford, 2003).

Marvin was skeptical of monetary policymakers' ability to manage expectations in this way, especially since the theoretically optimal policies are typically time-inconsistent and thus may not be credible. His preferred approaches instead involved concrete policy actions or institutional changes whose effectiveness does not depend on convincing the public and markets that the central bank's future policy strategy has changed.

Marvin's 2000 paper proposed several specific policies that could loosen the constraint of the ZLB. An idea that particularly appealed to him — he would reconsider related issues in detail in his Jackson Hole paper 16 years later — was for the central bank to impose a variable charge on bank reserves to create what would effectively be a negative short-term interest rate on that asset. Following the imposition of a charge, banks' efforts to substitute into alternative liquid assets should lead the negative rate to spill over into money markets and perhaps to longer-term assets as well. Indeed, this approach would ultimately be successfully adopted by several central banks, including the European Central Bank and the Bank of Japan, and actively studied by others, including the Bank of England.

However, monetary theorist that he was, Marvin worried that the effectiveness of taxing bank reserves would be undermined by depositor arbitrage. If banks tried to pass on their negative returns to depositors, say by imposing new fees on checking accounts, people would have an incentive to hold cash instead of deposits, thus disintermediating the banking system. That arbitrage in turn would limit the central bank's ability to impose negative rates. In his 2000 and 2016 papers, Marvin proposed three possible solutions to this problem.

First, simply abolish cash (or large bills, which are easiest to store). That would solve the problem, presumably, but Marvin worried about the social costs — increased transactions costs for the unbanked, for example — of such a step. Incidentally, this solution would later be explored in detail by Rogoff (2016), who emphasized that the social costs would be balanced by important gains, including the reduction of money laundering and tax evasion.

Second, impose a carry tax on cash, allowing a negative return on currency at times when a negative rate was needed to support economic activity. In his 2000 paper, Marvin suggested that the carry tax be implemented through magnetic strips attached to bills, which at the time seemed technologically implausible. Today, the Federal Reserve is discussing the possibility of creating a central bank digital currency (CBDC). If a CBDC were to bear a variable interest rate, and if it supplanted physical currency, it would bring Marvin's idea into the realm of technical feasibility.

Finally, in his 2016 paper, Marvin suggested that the Fed eliminate the fixed one-for-one exchange rate between currency and bank deposits and instead allow the exchange rate to be determined by market conditions. By varying the supply of currency, the central bank would then be able to keep the return on currency close to the return on deposits during periods of negative rates. This approach sounds strange to modern ears, but there is historical precedent. Before the creation of the Fed, when bank runs led to the suspension of convertibility of deposits into cash, bank deposits and currency often traded at varying, market-determined rates.

One can only be impressed by the intellectual rigor and creativity of these ideas. With the benefit of hindsight, though, it does not seem that any of these additional measures are needed to achieve meaning-fully negative rates (though I take Rogoff's point about the negative social externalities caused by the circulation of large bills). If anything, Marvin underestimated the effectiveness of his proposal to impose a variable charge on bank reserves, without supplementary steps to limit currency hoarding. In recent years, central banks have been able to push beyond the ZLB without provoking the hoarding that Marvin worried about. In part, that reflects the fact that banks rely substantially on wholesale funding markets. The costs to lenders in those markets of holding assets in physical cash is substantial, taking into account security, insurance, and the inconvenience of making large payments to global counterparties in cash. Banks have thus been able to pass on negative returns to many of their largest funders, even if not to all retail depositors.

Also with the benefit of hindsight, two other issues would have to be addressed before adopting any of Marvin's proposals regarding currency. First, much of the recent discussion about negative rates has centered on possible risks to financial stability. If significantly negative rates make banks, money market funds, or other institutions unprofitable or unstable, it is possible that there is a "reversal" interest rate below which rate cuts are no longer beneficial.2 Neither of Marvin's papers mention this issue. Personally, I am less alarmed than some about this risk: Negative rates, at the levels we have seen them, appear to have eased broad financial conditions without creating serious problems for financial institutions, over and above the effects of low rates in general. But this issue is one that would require more study before rates were cut much more deeply.

The other barrier to the implementation of Marvin's proposals is the difficulty of gaining political support. Changes to the currency, even regarding (economically) trivial matters like whose face is on the bill, can be highly controversial. The Fed is independent, but Congress has ultimate control over the currency and will oppose unpopular measures. Here is my opportunity to praise Marvin's intellectual integrity. The proposal in his Woodstock, Vermont, paper to impose a carry tax on currency created a media firestorm, which was soon followed by a statement from the Libertarian Party and a proposed bill from Congressman Ron Paul (R-TX) to prohibit fees on currency. The Federal Reserve Board staff had to assure legislators that research by Fed economists is conducted independently and does not necessarily represent the views of the Federal Reserve System. I have been told (though cannot document) that Marvin received threats to his personal safety. Marvin believed in free intellectual inquiry and was not dissuaded. He made similar recommendations in the high-profile Jackson Hole meeting in 2016.

Even though Marvin's work on the ZLB focused on the constraint imposed by the availability of zero-interest cash, he looked at other possible approaches to overcoming the constraint, again always with thoughtful prescience. In his 2000 paper, beyond negative rates, Marvin anticipated another key tool used by all major central banks at the ZLB — namely, purchases of longer-term securities, or what today we call quantitative easing (QE).

Marvin's early exposition of how QE might work had a more monetarist flavor than most modern accounts. But he did not make the error of assuming that adding liquidity — in the narrow sense of assets useful primarily as transactions media — when the economy is in a liquidity trap would have much stimulative power. He instead took a broader view of the liquidity services provided by longer-term securities such as Treasuries, including not only the ease with which they can be turned into cash, but their usefulness as collateral and in providing other services. He saw central bank purchases of longer-term securities as affecting the convenience yields of those securities (as reflected in term premiums) and their close substitutes, in a manner very similar to the portfolio balance channel emphasized by Bernanke (2020) and most contemporary discussions of QE. In particular, by raising the values of assets broadly, Marvin saw QE as raising borrower net worth and thus reducing the external finance premium — the wedge between the safe rate of return and the cost of borrowing — in the sense of Bernanke and Gertler (1995).

Does QE have costs as well as benefits? In his 2000 paper, Marvin focused primarily on the fiscal risks of the central bank holding long bonds. The possibility of losses on the central bank's portfolio is indeed a concern for monetary policymakers but mostly for optical and political reasons. Losses on securities bought in QE programs are not first-order social costs, any more than are the losses that arise when the maturity mix of new securities issued by the Treasury turns out not to have been cost-minimizing. Moreover, any paper losses on central bank portfolios should be set against the benefits of QE for economic growth and thus for tax revenues. Finally, central banks can continue to operate with losses or low levels of capital, so agreements with the Treasury to replenish central bank capital — as we have seen in the United Kingdom, for example — relate more to questions of governance and independence rather than to the feasibility of QE policies.

The fiscal and political concerns about QE are shared by many central bankers. More puzzling is Marvin's view, stated in 2000 and reiterated in his 2016 paper, that QE risks becoming "inflationary finance." In the depressed conditions that followed the global financial crisis, and with the ZLB binding, the use of QE proved insufficient to get inflation even up to target, much less to uncomfortable levels. In particular, with rates very low, the velocity of money fell significantly during the post-crisis period. Marvin was perhaps concerned that QE would not be slowed or reversed after the economy no longer required support. It is true, at least, that unwinding a QE program without disrupting markets too much can take some time, during which the central bank's securities holdings continue to provide stimulus.

Finally, Marvin's 2000 paper considered yet another strategy for defeating the ZLB — so-called helicopter money, Milton Friedman's term for money-financed tax cuts. Helicopter money — or monetary transfers, in Marvin's language — is a combination of tax cuts and QE, a combination we have seen in many countries during the recent pandemic. We have good reasons to expect that transfers and tax cuts, on the one hand, and central bank securities purchases, on the other, will stimulate the economy. The question is whether the combination of the two policies is any more powerful than the policies taken separately, that is, whether monetary-fiscal coordination buys anything in this instance. Marvin appreciated in 2000, very early in this debate, that the answer may be no. Unless the public sees monetary-fiscal coordination as changing the goals of the central bank in the medium term — for example, by increasing the amount of inflation it is willing to accept — the stimulus provided by the combined policies will be roughly the same as that of the two policies taken separately. For helicopter money to have extra stimulative power, there must be either a real or perceived increase in fiscal dominance (loss of central bank independence), which the public believes will lead the central bank to accept higher inflation than it otherwise would, at least for a time.

The effects of the ZLB, and strategies for overcoming it, was just one of many critical issues in monetary theory and policy that Marvin Goodfriend tackled in his career. He was an extraordinary economist. He was also an extraordinarily kind person, with the gift of telling you that you are completely wrong while making you feel good about it. I, among his many friends, will miss him.


Bernanke, Ben. 2020. "The New Tools of Monetary Policy." American Economic Review 110, no. 4 (April): 943-83.

Bernanke, Ben, and Mark Gertler. 1995. "Inside the Black Box: The Credit Channel of Monetary Policy Transmission." Journal of Economic Perspectives 9, no. 4 (Fall): 27-48.

Brunnermeier, Markus, and Yann Koby. 2018. "The Reversal Interest Rate." NBER Working Paper 25406, December.

Eggertsson, Gauti, and Michael Woodford. 2003. "The Zero Bound on Interest Rates and Optimal Monetary Policy." Brookings Papers on Economic Activity no. 1 (Spring): 139-211.

Goodfriend, Marvin. 2000. "Overcoming the Zero Bound on Interest Rate Policy." Journal of Money, Credit and Banking 32, no. 4, part 2 (November): 1007-35.

Goodfriend, Marvin. 2016. "The Case for Unencumbering Interest Rate Policy at the Zero Bound." Jackson Hole Policy Symposium, Federal Reserve Bank of Kansas City, August 26-27.

Krugman, Paul. 1998. "It's Baaack: Japan's Slump and the Return of the Liquidity Trap." Brookings Papers on Economic Activity no. 2 (Fall) 137-87.

Rogoff, Kenneth. 2016. The Curse of Cash: How Large-Denomination Bills Aid Crime and Tax Evasion and Constrain Monetary Policy. Princeton, N.J.: Princeton University Press.

Wolman, Alexander. 1998. "Staggered Price Setting and the Zero Bound on Nominal Interest Rates." Federal Reserve Bank of Richmond Economic Quarterly 84 (Fall), 1-24.

Cite as: Bernanke, Ben S. 2022. "Marvin Goodfriend and the Zero Lower Bound." In Essays in Honor of Marvin Goodfriend: Economist and Central Banker, edited by Robert G. King and Alexander L. Wolman. Richmond, Va.: Federal Reserve Bank of Richmond.


With the advent of modestly negative rates in some jurisdictions, the lower bound is now commonly referred to as the effective lower bound. Here I'll stick with the older ZLB terminology.


Brunnermeier and Koby (2018).

Phone Icon Contact Us

Lisa Davis (804) 697-8179